The Evolution and Outflows of Hyper-Accreting Disks Brian Metzger, UC Berkeley with Tony Piro, Eliot Quataert & Todd Thompson Metzger, Thompson & Quataert (2007), ApJ, 659, 561 Metzger, Quataert & Thompson (2008), MNRAS, 385, 1455 Metzger, Thompson & Quataert (2008), ApJ, 676, 1130 Metzger, Piro & Quataert (2008a), MNRAS in press Metzger, Piro & Quataert (2008b), In preparation Outline Introduction Compact Object Mergers and White Dwarf AIC Short GRBs: Recent Advances and New Puzzles Hyper-Accreting Disk Models One-Zone “Ring” Model 1D Height-Integrated Model Disk Outflows and Nucleosynthesis Neutrino-Driven Winds (Early Times) Viscously-Driven Winds (Late Times) Conclusions Compact Object Mergers (NS-NS or BH-NS) Lattimer & Schramm 1974, 1976; Paczynski 1986; Eichler et al. 1989 t = 0.7 ms “Chirp” Shibata & Taniguchi 2006 t = 3 ms • Inspiral + NS Tidal Disruption – Primary Target for Advanced LIGO / VIRGO • • • • Disk Forms w/ Mass ~ 10-3 - 0.3 M and Radius ~10-100 km Hot ( kT > MeV) and Dense ( ~ 108-1012 g cm-3) Midplane Cooling via Neutrinos: ( >>1, ~ 0.01-100 ) Ý ~ 102 10M s-1 GRB Progenitor? Accretion Rate M Accretion-Induced Collapse Nomoto & Kondo 1991; Canal 1997 • Electron Capture (24Mg 20Ne 20O) Faster than Nuclear Burning O-Ne-Mg White Dwarf Core Destabilized QuickTime™ and a TIFF (LZW) decompressor are needed to see this picture. Dessart+06 776 ms post bounce Md ~ 0.1 M Disk Forms Around NS Gamma-Ray Bursts: Long & Short Duration BATSE GRBs Long • High Redshift: <z> ~ 2 • Large Energies (Eiso~1052-54 ergs) • Star Forming Hosts • Type Ibc Broad-Line Supernovae Nakar 07 Short 2 1 H t visc K R 1/ 2 M BH 0.1 R 3 / 2 H /R 2 0.4 s 100 km 0.2 3M Short GRB Host Galaxies GRB050709 GRB050509b Bloom+ 06 z = 0.16 SFR = 0.2 M yr-1 z = 0.225 SFR < 0.1 M yr-1 KECK Bloom+06 HUBBLE Fox+05 GRB050724 Berger +05 z = 0.258 SFR < 0.03 M yr-1 Berger+05 Short GRB Host Galaxies GRB050709 GRB050509b Bloom +06 z = 0.16 SFR = 0.2 M yr-1 No SN! (But Some Radioactive Ejecta Expected…) z = 0.225 SFR < 0.1 M yr-1 KECK Bloom+06 HUBBLE Fox+05 • Lower z • Eiso~ 1049-51GRB050724 ergs Berger +05• Older Progenitor GRB050724 Population z = 0.258 SFR < 0.03 M yr-1 Berger+05 Short GRBs with Extended Emission GRB050709 Who Ordered That?! - Regular ~ 30-100 s Duration - Energy Often Exceeds GRB’s - ~25% of Swift Short Bursts BATSE Examples Late-Time Flaring (Norris & Bonnell 2006) GRB050724 XRT, Campana+06 A “Ring” Model of Hyper-Accreting Disks Metzger, Piro & Quataert 2008a MÝ BH Vr < 0 Vr > 0 rd Ý • Mass at large radii ~ rd controls disk evolution and sets M • Model enforces mass & angular momentum conservation • Thermal Balance: dS T qÝvisc qÝ dt qÝvisc 9 2 , 4 Md tvisc (rd ) = cs H • Calculates {, T, H} @ rd(t) GIVEN rd,0, Md,0, MBH, and Simple model allows wide exploration of parameter space: Initial disk mass/radius, viscosity , outflows, etc. Three Phases of Hyper-Accreting Disks 2 3 1 Ý Thick Disk: H ~ R 1) High M - Optically Thick Matter Accretes Before Cooling 2) Neutrino-Cooled Thin Disk: H ~ 0.2 R - Ý c2 Optically Thin, Neutrino Luminosity L ~ 0.1 M Ion Pressure Dominated / Mildly Degenerate e p n e Neutron-Rich Composition (n/p ~ 10) Ý Thick Disk: H ~ R 3) Low M - e n e p Neutrino Cooling << Viscous Heating Radiation Pressure-Dominated / Non-Degenerate Example Ring Model Solution . 0.1 Mdc2 (1051 ergs) T (MeV) rd (km) M (M s-1) MBH = 3 M Md,0 = 0.1 M rd,0 = 30 km = 0.1 tvisc,0 ~ 3 ms Mdt-1/3 tthick Late-Time Thick Disk Outflows Advective disks are only marginally bound. When the disk cannot cool, a powerful viscously-driven outflow blows it apart (Blandford & Begelman 1999). Only a small fraction of ingoing matter actually accretes onto black hole BH Hawley & Balbus 2002 QuickTime™ and a YUV420 codec decompressor are needed to see this picture. Nuclear energy from -particle formation also sufficient to unbind disk Effect of the Thick Disk Wind Late-Time Short GRB Activity tthick? • XRBs Make Radio Jets Upon Thermal (Thin Disk) Power-Law (Thick Disk) Transition (e.g. Fender +99; Corbel + 00; Fender, Belloni, & Gallo 04; Gallo +04) • Extended Emission = Thick Disk Transition? • Problem: Requires Very Low Viscosity ~ 10-3 23/17 tTHICK ~ 0.1 s 0.1 M d ,0 9 /17 rd ,0 9 / 34 M BH 1/ 2 0.1M 100 km 3M Other Sources of Extended Emission Lee & Ramirez-Ruiz 07 Tidal Tail Fallback Rosswog 06, Lee & Ramirez-Ruiz 07 Magnetar Spin-Down Following AIC P0= 1 ms 1016 G GRB060614 Overlaid Metzger, Quataert & Thompson 08 NS EÝ Ýc 2 M High 3 1015 G Low 1015 G Disk Outflows & Heavy Element Synthesis • GRB Jets Require Low Density, but High Density Outflows Probably More Common Heavy Element Formation {n, p} ' s C Fe - group r - process? 12 • EBIND ~ 8 MeV nucleon-1 vOUT ~ 0.1-0.2 c • Which Heavy Isotopes are Produced Depends on: Electron Fraction Ye = np/(nn+np) Ye Product Nuclei 0.48 - 0.6 Mostly Ni56 - Ideal 9 Day Decay Time 0.4 - 0.48 Rare Neutron-Rich Isotopes (58Fe, 54Cr, 50Ti, 60Zn) 0.3 - 0.4 Very Rare Neutron-Rich Isotopes (78,80,82Se, 79Br) < 0.3 r-Process Elements (e.g. Ag, Pt, Eu) (Ye = 0.88) (Ye ~ 0.5) Rare Neutron-Rich Isotopes (Ye ~ 0.3 - 0.4) 2nd/3rd Peak r-Process (Ye < 0.3) (Ye < 0.2) Atomic Number (A) Neutrino Heated Winds Burrows, Hayes, & Fryxell 1995 Original Application: Core-Collapse Supernovae (Duncan+ 84; Qian & Woosley 96; Thompson+ 01) t = 0.5 s Emergence of the Proto-Neutron Star Wind e n e p p e n e • Neutrinos Heat & Unbind Matter from NS: • Electron Fraction at set by Neutrinos n p – EBIND = 150 MeV, E ~ 15 MeV ~ 10 Neutrino Absorptions per Nucleon NÝe e n Le Ee n Ý p N e e p L e E e n p 1 Le Ee Ye Ye 1 L E e e n Neutrino-Driven Accretion Disk Winds Levinson 06; Metzger, Thompson & Quataert 08 GMmn IF E THEN Ye Ye 2R GMmn disk IF E THEN Ye Ye 2R Ýc2 L ~ 0.1 M Ý M Wind (R), Ye Yedisk ~ 0.1 BH Neutrino Luminosities L e /L e and Mean Energies E e /Ee Ýdisk Calculated Using a Steady - State Disk Model Given M 56Ni Production in Neutrino-Driven Winds 1 rd 56Ni 10-1 GMmp/2R < E Optically Thin @ RISCO GMmp/2R > E Accretion Rate (M s-1) Optically Thick @ RISCO Neutron-Rich Isotopes 10-2 1 10 Wind Launching Radius (RISCO) Metzger, Piro & Quatert 2008 Neutron-Rich Isotopes Mini-Supernovae Following Short GRBs Li & Paczynski 1998; Kulkarni 2005; Metzger, Piro & Quataert 2008a Total 56Ni Mass Integrated Over Disk Evolution: (MNi ~ 10-3 M and Mtot ~ 10-2 M) GRB050509b (Hjorth +05) V J Optical / IR Follow-Up Initial Disk Properties Metzger, Piro & Quataert 2008a Metzger, Piro & Quataert 2008a BH spin a = 0.9 Mini-SN Light Curve Summary So Far Neutrino-Cooled Thin Disk Phase - Neutron-Rich Midplane (Ye ~ 0.1) - Neutrino-Driven Wind Up To ~ 10-3 M in 56Ni Mini-SN (+ even more neutron-rich matter from larger radii) Late-Time Thick Disk Phase - Viscously-Driven Wind Disrupts Disk - Disk Composition?? Wind Composition?? Late-Time Disk Composition: Metzger, Piro & Quataert 2008b Disk Thickening Weak Freeze-Out The Thick Disk Transition Degeneracy Pair Captures: e p e n e n e p H/R Yeeq Ye Both Cool Disk AND Change Ye Weak Freeze Out Non-Degenerate Transition Moderately Neutron-Rich Freeze-Out (Ye ~ 0.25 - 0.45) 1D Height-Integrated Disk Calculations Equations Md,0 = 0.1 M, rd,0 = 30 km, = 0.3 Local Disk Mass r2 (M) Angular Momentum / Continuity Entrop y Heating Cooling Nuclear Composition QuickTime™ and a YUV420 codec decompressor are needed to see this picture. Weak Freeze-Out (A “Little Bang”) Weak Interactions Drive Ye Yeeq Until Freeze-Out Electron Fraction Ye Yeeq QuickTime™ and a YUV420 codec decompressor are needed to see this picture. Thickening / Freeze-Out Begins at the Outer Disk and Moves Inwards Neutron-Rich Freeze-Out Is Robust M per bin M0 = 0.1 M, = 0.3 Mtot = 0.02 M M0 = 0.1 M, = 0.03 Mtot = 0.02 M M per bin M0 = 0.01 M, = 0.3 Mtot= 2 10-3 M ~10 - 30% of Initial Disk Ejected Into ISM with Ye ~ 0.2-0.4 Production of Rare Neutron-Rich Isotopes Hartmann +85 40 Million Times Solar Abundance!!! 0.35 < Ye < 0.4 78,80,82Se, 79Br Ye = 0.5 =1-2Ye Ye = 0.4 Ye = 0.35 Merger Rates and the Short GRB Beaming Fraction Metzger, Piro & Quataert 2008b 1 M d ,0 1 5 1 NÝmax ~ 10 yr galaxy 0.2 0.1M 1 From known merging NS systems, Kim+06 estimate: NÝNSNS = 3 105 2 104 yr -1 Milky Way Short GRB Rate ~ 10-6 yr-1 (Nakar 07) M d ,0 NÝSGRB fb 0.13 Ý 0.2 0.1M N max Jet Opening Angle > 300 Short GRBs Less Collimated than Long GRBs (LGRB~2-100) (Grupe +06; Soderberg +06) Timeline of Compact Object Mergers 1) 2) 3) Inspiral, Tidal Disruption & Disk Formation (t ~ ms) Optically-Thick, Geometrically-Thick Disk (t ~ ms) Geometrically-Thin Neutrino-Cooled Disk (t ~0.1-1 s) - Up to ~ 10-3 M in 56Ni from neutrino-driven winds (mini-SN) 4) Radiatively Inefficient Thick Disk (t > 0.1-1 s) - Degenerate Non-Degenerate - PGAS-Dominated PRAD-Dominated - Neutron-Rich Freeze-Out Disk Blown Apart by Viscously-Driven Outflow - Creation of Rare Neutron-Rich Elements (“Little Bang”) From AIC Disk Winds Neutrino absorptions don’t affect Ye strongly in compact merger disks BUT In AIC, e “flash” from shock break-out can drive Ye > 0.5 e n e p Freeze-Out Ye in AIC Disk With e Flash Neutrino Luminosity (ergs s-1) 56Ni Dessart+ 06 L e “Flash” Le Time After Core Bounce (s) Winds synthesize ~10-2 M in 56Ni Optical Transient Surveys: ~ few yr-1 Pan-STARRs & PTF No e Flash ~ 100’s yr-1 LSST Neutron-rich material also synthesized? unusual spectral lines? (e.g, Zn, Ge, Cu?) Conclusions Isolated Disk Evolution Cannot Explain Late-Time X-Ray Emission (unless ~ 10-3) Promising alternatives: Tidal tail fall-back and magnetar spin-down Neutrino-driven winds create up to ~10-3M in 56Ni Mini-SN at t ~ 1 day Neutron-Rich Nucleosynthesis CO merger rate: < 10-5 yr-1 (Md,0/0.1 M)-1 Short GRB jet opening angle: > 30(Md,0/0.1 M)1/2 ~10-2 M in 56Ni from White Dwarf AIC Target for upcoming optical transient surveys Future Progress Observations Gravitational Waves (LIGO; VIRGO) Short GRB Optical / IR Follow-Up Spectroscopy of MetalPoor Halo Stars Optical Transient Surveys Theory MHD Disk Simulations: Freeze-Out and LateTime Winds Neutron-Rich Nucleosynthesis Compact Object Merger Simulations Spectra of Neutron-Rich Explosions Late-Time Optical Rebrightening: Mini-Supernova? GRB060614; Mangano+07 Merger Rates and the GRB Beaming Fraction • If a fraction ~ 0.1 of initial disk mass is ejected with Ye < 0.4 per event: NÝ M d,0 X t galaxy X For tgalaxy = 10 Gyr and MISM = 109 M: M ISM From known merging NS systems, Kim+06 estimate: NÝNSNS = 3 105 2 104 yr -1 Milky Way Short GRB Rate ~ 10-6 yr-1 (Nakar 07) Jet Opening Angle > 100 Short GRBs Less Collimated than Long GRBs (LGRB~2-100)